Could somebody help me out with a few problems I have implementing the
Bidirectional Algorithm as described in Unicode Technical Report #9
(www.unicode.org/unicode/reports/tr9/).
At one point in the report it states that the sequence:
DID YOU SAY 'he said "car MEANS CAR"'?
should be displayed as:
?'he said "RAC SNAEM car"' YAS UOY DID
My implementation displays the following:
?'"he said "RAC SNAEM car' YAS UOY DID
Part of the problem I have is that I don't know where the explicit
directional embeddings are being made if any. Currently I am assuming
that the sequence to be displayed is encoded as:
DID YOU SAY 'he said "<RLE>car MEANS CAR<PDF>"'?
which results in the following embedding level resolutions (my
implementation above with report's below):
DID YOU SAY 'he said "car MEANS CAR"'?
11111111111112222222224443333333333111 (Mine)
11111111111112222222224443333333333211 (report)
The difference is at the last quotation mark. I know it makes sense to
have that character at level 2 but how is it possible? Is there another
directional embedding before the "h"? An LRE? That would probably work
but am I making the right assumptions?
And does anybody have any idea what I'm doing wrong if there are no
embeddings there whatsoever?
Also, the report mentions implicit directional marks RLM, and LRM. Do
these have assigned Unicode values? How do I use them?
Thanks for your time,
Dave.
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:56 EDT